Project Summary/Abstract The purpose of this proposed study is to develop and validate threading framework for accurately predicting full-length helical trans-membrane proteins; then apply these algorithms to predict the structures of human G-protein coupled receptor (GPCR) proteins and their spliced variants. Trans-membrane proteins are very important aspect of human biological function, they serve critical roles in cellular processes such as respiration, signal transduction, cell trafficking, and transport of compounds and ions across cellular membranes. Dysfunctions in some trans-membrane proteins have been associated with diseases such as Alzheimers and diabetes [1, 2]. They are also a target for more than 50% of pharmaceutical drugs. To better understand the mechanism of their function, which in turn help better understand the mechanism of their associated diseases, the three dimensional (3D) structure must be determined. This research seeks to provide a threading framework for predict the 3D structure of membrane proteins, using GPCR as case study, through the technique called threading. Threading is when a query sequence is aligned to all representative structures energetically. To achieve the above aims, a threading-based framework for predicting the backbone structures of alpha helical trans-membrane proteins will be developed, then adapt exiting structural refinement tools to obtain full-atom structures. Next, the threading framework obtained and refined in step one and two will be applied to the about 701 human rhodopsin GPCR and its spliced variants.